Introduction
Salivary gland tumours are rare and constitute approximately 3-6% of all head and neck neoplasms. Most lesions are benign and located in the parotid gland. The preoperative differentiation between benign and malignant lesions is of crucial importance for patient management. However, the radiological findings, as well as fine needle aspiration cytology, may considerably overlap. This study aims to develop and evaluate a deep learning network for characterizing salivary gland tumours, based on non-contrast CT images.
Materials and methods
Pre-operative non-contrast CT volumes of 88 salivary gland tumours were included. After the 3D segmentation of the lesion with a modified V-Net model, ResNet-50 was trained to classify benign and malignant lesions as selected 2D slices of the region. The training and test sets consisted of 61 and 27 cases, respectively.
Results
Regarding the segmentation step, our methodology obtained on the test set the Dice score 0.85, and the 95% quantile-Hausdorff distance 4.6 on average. For the final step of the classification, the obtained accuracy is 0.89 and the F1-score 0.88 on average.
Conclusion
The proposed model has proven promising for salivary gland tumour characterization. This might improve patient management and aid the surgical strategy making a more accurate preoperative lesion classification.

Deep learning for salivary gland tumours segmentation and classification based on CT images

This presentation discusses the concept of quality in radiomics research and how to assess it, providing a brief overview of the current state of the art.

Assessing research quality in radiomics and AI

An increasing number of AI solutions is being developed and made available for implementation in clinical practice and radiology. All steps of both the development and implementation of such solutions need to be carefully evaluated however, and one needs to be aware of the potential risks and problems that can be associated with using such applications. In this presentation a stepwise approach is explained for addressing these issues. 

AI from bench to bedside

Background: The rising attention of MRI-based radiomics in nasopharyngeal carcinoma (NPC), have been shown not only among radiologists but also among oncologists. Aim of the review is to systematically evaluate the methodological quality of MRI-based radiomic studies of NPC and analysis of the publication journal characteristics. Methods: A systematic search was performed and the methodological quality of study included (=24) was evaluated according to the RQS (Radiomic quality score). Subgroup, for journal type (imaging/clinical) and biomarker (prognostic/predictive), and correlation, between RQS and journal Impact Factor(IF) and journal quartile and between IF and journal type/biomarker, were performed. Mann-Whitney-U test and Spearman’s correlation were performed. Results: The mean RQS was 7.5 ± 5.4. No studies reported a phantom study or a test re-test, biological correlation or open science data. No significant differences of RQS were detected between articles published in clinical/imaging journal and between studies with a predictive or prognostic biomarker as well as between total RQS and IF of the year of publication and journal quartile. A significant higher IF was found in radiomic studies published in clinical journal (p always < 0.05). Conclusions: Radiomic articles in NPC are mostly of low methodological quality, however they are of interest for clinicians.

The use of Radiomic Quality Score (RQS) assessment in MRI based radiomics in nasopharyngeal cancer: systematic review and perspective.

Objectives
The goal of the present study was to classify the most common types of plain radiographs using a neural network and to validate the network’s performance on internal and external data. Such a network could help improve various radiological workflows.
Methods
All radiographs from the year 2017 (n = 71,274) acquired at our institution were retrieved from the PACS. The 30 largest categories (n = 58,219, 81.7% of all radiographs performed in 2017) were used to develop and validate a neural network (MobileNet v1.0) using transfer learning. Image categories were extracted from DICOM metadata (study and image description) and mapped to the WHO manual of diagnostic imaging. As an independent, external validation set, we used images from other institutions that had been stored in our PACS (n = 5324).
Results
In the internal validation, the overall accuracy of the model was 90.3% (95%CI: 89.2–91.3%), whereas, for the external validation set, the overall accuracy was 94.0% (95%CI: 93.3–94.6%).
Conclusions
Using data from one single institution, we were able to classify the most common categories of radiographs with a neural network.

Practical applications of deep learning: classifying the most common categories of plain radiographs in a PACS using a neural network

Avascular necrosis and transient osteoporosis of the hip are conditions related to bone marrow edema of the proximal femur. Differentiation between the two entities can be extremely complicated, requiring significant MSK radiology expertise and the combination of imaging and clinical data. The aim of this study was to employ radiomics for the differentiation of the two entities with the use of MRI data. A total of 109 hips with TOH and 104 hips with AVN were retrospectively included and femoral heads were manually segmented. Radiomics features were extracted with the use of PyRadiomics as implemented in 3D Slicer. Relevant radiomics features (n=38) were selected by removing collinearities and applying Boruta tree-based feature selection. An extreme gradient boosting (XGboost) machine learning model was trained on 70% and validated on 30% of the dataset and the results were compared to the performance of two fellowship-trained MSK radiologists. XGboost achieved an Area Under the Curve (AUC) of 93.7% (95%CI from 87.7 to 99.8%) whereas the MSK radiologists achieved an AUC of 90.6% (95%CI from 86.7% to 94.5%) and 88.3% (95%CI from 84% to 92.7%) respectively. In conclusion, radiomics-based machine learning achieved excellent performance, similar to MSK radiologists, in differentiating between TOH and AVN.

Radiomics analysis enables accurate differential diagnosis between avascular necrosis and transient osteoporosis of the hip

With recent developments, an extensive and increasing amount of medical imaging data is produced every day. This data provides an opportunity for researchers to develop data-driven methods (using e.g. deep learning) and deliver better healthcare. However, data-driven models require large amounts of data to be properly trained while there is always a limited amount of data available at each center. Hence, deep learning models trained on only local data cannot reach their full performance capacity. One solution could be to accumulate all data from different centers into one center. However, data privacy regulations do not allow medical institutions to combine their data easily. Another solution is to use privacy-preserving algorithms, that use all data available in multiple centers while keeping the sensitive data private. Federated learning is such a mechanism that enables deploying large-scale machine learning models trained on different data centers without sharing sensitive data. In this mechanism, instead of transferring data, a general model is trained on local datasets and transferred between data centers. Federated learning is a promising field of research, with extensive use in medical research. This presentation introduces federated learning to radiologists, with an extensive look into its concepts and recent research trends in radiology.

Federated learning in radiology: challenges and promises

Public radiology report datasets are hard to obtain due to privacy issues, hindering the development of natural language processing (NLP) machine learning models. The scarcity of radiology reports datasets, especially in Portuguese, is a major drawback in the advancement and development of NLP models. We present here a public dataset called the "UNIFESP Rad Report Dataset v1.A'' that will be continually updated. The construction of the dataset was done by extracting every CT scan radiology report from 2014-2021. We performed automatic anonymization using Regex to remove patient and physician names, identification numbers, and dates. The second step of anonymization was listing unique words and performing a manual replacement of them in the reports. The last step was during the annotation process we searched for any remaining identification in the report that was not removed by our automated process and we did manual removal. The first version of the dataset comprises 557 de-identified radiology reports of CT scans from different body parts and annotations for critical findings (74 positives, 483 negatives). The dataset is available at https://github.com/DDI-UNIFESP-AI-Informatics-in-Radiology/UNIFESP-Radiology-Report-Dataset and will be constantly updated.

The Unifesp Radiology Report Dataset

Introduction Artificial intelligence (AI) has started to be increasingly adopted in medical imaging and radiotherapy clinical practice, however research, education and partnerships have not really caught up yet to facilitate a safe and effective transition. The aim of the document is to provide baseline guidance for radiographers working in the field of AI in education, research, clinical practice and stakeholder partnerships. The guideline is intended for use by the multi-professional clinical imaging and radiotherapy teams, including all staff, volunteers, students and learners. Methods The format mirrored similar publications from other SCoR working groups in the past. The recommendations have been subject to a rapid period of peer, professional and patient assessment and review. Feedback was sought from a range of SoR members and advisory groups, as well as from the SoR director of professional policy, as well as from external experts. Amendments were then made in line with feedback received and a final consensus was reached. Results AI is an innovative tool radiographers will need to engage with to ensure a safe and efficient clinical service in imaging and radiotherapy. Educational provisions will need to be proportionately adjusted by Higher Education Institutions (HEIs) to offer the necessary knowledge, skills and competences for diagnostic and therapeutic radiographers, to enable them to navigate a future where AI will be central to patient diagnosis and treatment pathways. Radiography-led research in AI should address key clinical challenges and enable radiographers co-design, implement and validate AI solutions. Partnerships are key in ensuring the contribution of radiographers is integrated into healthcare AI ecosystems for the benefit of the patients and service users. Conclusion Radiography is starting to work towards a future with AI-enabled healthcare. This guidance offers some recommendations for different areas of radiography practice. There is a need to update our educational curricula, rethink our research priorities, forge new strong clinical-academic-industry partnerships to optimise clinical practice. Specific recommendations in relation to clinical practice, education, research and the forging of partnerships with key stakeholders are discussed, with potential impact on policy and practice in all these domains. These recommendations aim to serve as baseline guidance for UK radiographers. Implications for practice This review offers the most up-to-date recommendations for clinical practitioners, researchers, academics and service users of clinical imaging and therapeutic radiography services. Radiography practice, education and research must gradually adjust to AI-enabled healthcare systems to ensure gains of AI technologies are maximised and challenges and risks are minimised. This guidance will need to be updated regularly given the fast-changing pace of AI development and innovation. 

Artificial intelligence: guidance for clinical imaging and therapeutic radiography workforce professionals , a society and college of Radiographers publication

Purpose
To develop a robust algorithm for classification of musculoskeletal x-ray images in the most common projections of predefined body regions, and to evaluate the performance of the algorithm. Classifying radiographs is important for applying region-specific algorithms. As there is currently no large data set with x-ray projections, one had to be created first. Our goal was also to gain insight into the decision of the network.
Methods
Musculoskeletal radiographs from our department from 2018-2019 were classified into predefined body parts and projections. The dataset consisted of anonymized studies from adult patients of all genders with various pathologies from which the most common 15 body parts and the most common 30 radiographic projections were selected. 14100 images were annotated by a medical student, followed by a junior and a senior radiologist. Data leakage was excluded by separating the data groups on a patient level. First a group of patients containing 500 images were randomly selected and separated for the network testing. The remaining patients were split 70:30 resulting in 9492 images for the training dataset and 4108 for the validation dataset. Data annotation was performed using our scientific medical imaging platform Nora. As a Deep Learning framework Tensorflow 2.4.0 with the included Keras framework was used. For robust results an established convolutional neural network, Inception-v3 by C. Szegedy et al., was selected. As Inception-v3 was optimized for 1000 classes the network top layer was removed and a new network top was added. It consists of a pooling layer, a dropout layer for preventing fast overfitting and a dense layer for selecting the output of the 30 classes. As X-ray images differ from Imagenet images we decided to leave the whole network fully retrainable and used the Imagenet weights as an optimized starting point for the base network. The input images were optimized as two color-channels would not be used by a grayscale image. These were filled using the inverted image and an edge-optimized image. The network input size was set to 256*256 pixels and the X-ray images were accordingly resized. On each training and validation step the corresponding datasets were randomized, homogeneously balanced, and mildly augmented. For augmentation a possible vertical flip, a mild random cropping and a small degree of random rotation and contrast changes were applied. The amount of training epochs was set to 200, using a batch size of 50 with 100 steps of batching per epoch. The initial learning rate was set to 0.1 and decayed steadily during training to 0.05. Training was performed with a server graphics unit (Nvidia Tesla K80). Scikit-learn was used to calculate the data statistics. For visualization the software package to tf-explain was used, especially the implementation of Gradient-weighted Class Activation Mapping (Grad-CAM). It uses the gradients of the final convolutional layer to produce a heat map of the impo

Deep learning for classification of musculoskeletal x-ray images.

Objective: Research in Artificial Intelligence (AI) for medical imaging requires large volumes of high-quality, labelled data. The Cancer Imaging Archive (TCIA) is a public repository of DICOM images related to oncology. The aim of this work is to provide researchers and developers with a simple way to import images from the TCIA servers onto a local PACS environment.
Methods: TCIA provides a network interface (REST API) that enables third-party applications to programmatically access its collections. Orthanc is an open-source DICOM server that can be deployed by research teams as their PACS server. The deliverable of this work is an original, open-source plugin for Orthanc that imports images from TCIA using its REST API.
Results: The developed plugin takes the form of an easy-to-use Web application to browse TCIA collections and import their images. The imported images are served according to the DICOM standard, and can be displayed using zero-footprint viewers.
Conclusion: This work proposes a solution to import real-world, open-data medical images into an environment that mimics clinical setups, which is essential to AI research. Those images come from TCIA, that contains multiple acquisitions of various body parts acquired by different modalities, making it useful in many fields of radiology.

Importing and serving open-data medical images to support Artificial Intelligence research

Objective: Research in medical imaging necessitates rigorous management of the image databases. Typical clinical trials and the training of AI algorithms require to manage thousands of DICOM instances. The aim of this work is to provide researchers with a standardized way to transparently index the content of a filesystem containing large amounts of heterogeneous imaging data.
Methods: The DICOM standard specifies the well-known patient/study/series/instance hierarchy as a model of the real world. Orthanc is an open-source DICOM server that can be deployed by research teams as their PACS server. We introduce a strategy to use Orthanc as a platform that publishes the content of a filesystem according to the DICOM hierarchy.
Results: The deliverable of this work is an original, open-source plugin for Orthanc that continuously synchronizes the content of an Orthanc server with the content of a filesystem. This way, the filesystem is automatically organized according to the DICOM model of the real world, without any manual intervention. The indexed DICOM resources are immediately available in a Web interface and in a Web viewer, and can be queried/retrieved by DICOM clients.
Conclusion: This work proposes a simple, automated method to publish filesystems containing medical images as a DICOM server.

Automatically publishing medical images from a filesystem as a DICOM server

This text is the summary of my presentation evaluating the current status of teleradiology applications and user satisfaction in Turkey in 2021. The survey was supported by TSR and consists of 24 questions. Data were collected through a web-based survey system. 153 radiologists answered the questionnaire in 3 months (March-May 2021). According to the results, the technical infrastructure and usage concepts of teleradiology in general are similar to global applications. Unfortunately, due to a lack of legislation, some of the practices have gotten out of control and have been used in unacceptable ways. Currently the main motivation for teleradiology is financial income. But the prices are very low and unreasonable. Low prices lead to poor reporting quality and dissatisfaction with teleradiology. While TSR helps radiologists improve teleradiology standards, an acceptable and efficient teleradiology service in favor of patients will only be possible with appropriate government regulations and controls.

Turkish survey of radiologists' experience, workloads, technical infrastructures and satisfactions on teleradiology practice. (Turkish Society of Radiology)

Purpose : Chest radiography is the first-line imaging modality for diagnosing thoracic pathologies. Diagnostic accuracy may be drastically reduced due to technical limitations resulting in poor image quality and thus leading to incorrect diagnosis.
Methods : We collected 4481 frontal Chest X-rays (CXRs), performed in the supine or standing position, labeled by 7 radiologists from the radiology department of Cochin University hospital (AP-HP, Paris,France). Frontal CXR quality criteria included full anatomical coverage, lack of rotation or scapula projection, deep inspiration and optimal exposure. A deep convolutional neural network was trained to classify CXRs as technically correct or incorrect. The model predictions were compared to a ground truth set (15%), labeled by 2 chest radiologist experts.
Results : There were no signifiant differences regarding age, sex, and classification for each criterion between the test, validation and training datasets (p>0,05).We first evaluated the inter-rater reliability and found good correlation for each criterion (Cohen’s kappa > 0,6). The model performance was evaluated and compared to that of each observer with expert annotation as ground truth. The model performance was close to the radiologists' performance, (accuracies 80-90% compared to 83-95% for radiologists) excepted for the rotation criteria.
Conclusion : The trained model to detect quality-deficient CXRs could be used by technologists in real-time to ensure high-quality images.

Development of an AI-based model for Chest X-ray quality assessment

This study aims at comparing the performances of routine radiological interpretation and artificial intelligence (AI) in the detection of skeletal lesions. We retrospectively evaluated all radiographs performed after a recent trauma during 3 consecutive months in an imaging group.  Each examination was analyzed by the AI and its results were compared to those of the radiologists’ reports. In case of discrepancy, the examination was reviewed by a senior skeletal radiologist to settle on the presence of fractures, dislocations, elbow effusions, and focal bone lesions (FBL).  A total of 4774 exams were included in the study.  Lesion-wise sensitivity was 73.7% for the radiologists’ reports vs. 98.1% for the AI (+24.4 points) for fracture detection, 63.3% vs. 89.9% (+26.6 points) for dislocation detection, 84.7% vs. 91.5% (+6.8 points) for elbow effusion detection, and 16.1% vs. 98.1% (+82 points) for FBL detection. The specificity of the radiologists’ reports was always 100% whereas AI specificity was 88%, 99.1%, 99.8%, 95.6% for fractures, dislocations, elbow effusions, and FBL respectively. The NPV was measured at 99.5% for fractures, 99.8% for dislocations, and 99.9% for elbow effusions and FBL.

Performances of a deep learning algorithm for the detection of fracture, dislocation, elbow joint effusion, focal bone lesions on trauma X-rays

Purpose
We have created an augmented reality application that enables the user to simulate a CT biopsy using a simple home computer with a webcam.
Due to the disruption of many of the educational activities in the current climate relating to COVID-19, this offers an opportunity for students and trainees to gain some practical skills and motivation in relation to surgery.
The aim of this study is to assess the effectiveness of using the Augmented Reality to simulate a CT biopsy on an agar phantom.

Methods
We invited 12 Radiology trainees to use the software during a training day. All participants were given an information leaflet and asked to fill in a consent form.
They were asked to fill in a feedback form which asked which was anonymised and stored in a database. Participants were asked to rate the software in terms of the accuracy in comparison to an actual CT guided biopsy from 1 to 5 (1 - not accurate, 5 - accurate representation).
Results
The results were positive with participants rating the simulation of the procedure, accuracy of the needle placement and the haptic feedback from the gelatinphantom as 3.8 out of 5, 3.5 out of 5 and 3.3 of 5 respectively.
Comments included that it was a "great experience" and "interesting idea" and "the needle placement and visualisation can be a little challenging on the software"

Conclusion
We are taking the results of the feedback and are planning to improve the software.
After this, we aim to re-evaluate the software on future training days to see if the changes made improve the efficacy of the training . Any successful solution could benefit other training centers.

The effectiveness of augmented reality for CT biopsy skill training

Purpose:
To evaluate the influence on the performance of an automated vessel detection algorithm, of CT dose and reconstruction in anthropomorphic phantom images.
Methods: A 3D-printed vessel phantom (material Visijet-EX200; 0.1-4.25mm radius range) inside a PMMA thorax-shaped holder was scanned [CT-thorax protocol; (CTDIvol=4.0-2.1-1.0-0.5-0.2mGy); Canon_Aquilion_Prism, 4 repetitions] and images reconstructed with filtered-back-projection (FBP-FC08), iterative (AIDR3De-FC08) and deep-learning (DL) (AiCE-lung-standard) methods.
Images were analyzed with an automated in-house graph-cuts-based method for pulmonary vessel extraction and quantification, measuring, for each radius, the median pixel value (MPV, Hounsfield units-HU) and inter-quartile-range of pixel values (IQR, noise measurement) together with the total volume of voxels identified as vessels (averaged over 4 acquisitions).
Results:
As an example, for 3mm-radius, (MVP±σ) were, for (CTDIvol=0.2-0.5-1.0-2.1-4.0 mGy): [FBP: (98±5HU)-(97±9HU)-(93±7HU)-(94±7HU)-(91±2HU)]; [iterative: (101±3HU)-(104±4HU)-(104±7HU)-(102±4HU)-(99±3HU)]; [DL-based: (106±5HU)-(102±10HU)-(106±8HU)-(107±11HU)-(103±6HU)].
The IQR decreased with increasing dose for all reconstructions: (IQR±σ; 3mm-radius) were, for (CTDIvol=0.2-0.5-1.0-2.1-4.0 mGy): [FBP: (120±6HU)-(72±5HU)-(44±3HU)-(34±4HU)-(37±10HU)]; [iterative: (62±8HU)-(42±2HU)-(33±9HU)-(27±4HU)-(25±4HU)]; [DL-based: (62±12HU)-(59±8HU)-(43±7HU)-(39±5HU)-(36±6HU)].
The average detected vessel tree volume (ml) varied with dose and reconstruction: [FBP: (7.14±0.02ml)-(5.66±0.03ml)-(5.39±0.01ml)-(5.33±0.05ml)-(5.24±0.07ml)]; [iterative: (4.22±0.06ml)-(4.91±0.10ml)-(5.19±0.04ml)-(5.41±0.08ml)-(5.47±0.05ml)]; [DL-based: (6.36±0.07ml)-(7.15±0.08ml)-(7.22±0.06ml)-(7.39±0.03ml)-(7.42±0.03ml)].
Conclusion:
Vessel detection output depended on reconstruction method (with more vessels detected with DL reconstruction) and dose level. 3D-printed anthropomorphic phantoms with known structures are useful to test objectively the performance of automated tools for clinical diagnosis.

Impact of deep learning reconstruction and CT dose on automatic lung vessel morphology characterization software: a 3D printed anthropomorphic phantom study

Tumor response to therapy in clinical trials is typically evaluated with RECIST, aiming to assess tumor burden objectively. However, even when correctly employed, RECIST evaluations still suffer from inconsistencies. This is primarily a result of the subjectivity in the selection of lesions at baseline, which is limited to a maximum of 5 lesions.
We conducted a mathematical simulation to analyze the factors contributing to RECIST variability, when the inter-reader measurement error is removed.
A patient model, parameterized by the number of lesions, organs, mean tumor growth and variance, was implemented. Multiple virtual cohorts of patients were created, by fixing all but one of the input parameters to a default value and varying the remaining within a range. Each patient was assessed by multiple virtual readers, who randomly select valid target lesions according to RECIST criteria. The discordance (D) in the selection of lesions by the different readers that resulted in divergent RECIST categories was computed.
The choice of lesions at baseline has a sizable impact on RECIST assessments. Variability aggravates when the number of measurable lesions is large (D up to 25%) and when lesion growth borders the thresholds for progressive disease and partial response (D up to 40%).

A radiological simulation computer model to analyze intrinsic RECIST variability

Cancer survival prediction could potentially support individualized treatment planning of cancer patients. Healthcare centers accumulate several types of data that could be leveraged by artificial intelligence to create multimodal prognostic models. This study investigates the potential benefits of integrating radiological CT imaging of the thorax, laboratory blood tests, and clinical data to predict the 1-year survival in a pancancer cohort of non-small cell lung cancer, melanoma, and bladder cancer patients treated with immunotherapy. The study included 447 patients with longitudinal CT scans and 475 patients with longitudinal lab data. A 3D variant of the ResNet convolutional neural network was trained with imaging, and two random forest models were trained with lab and clinical data to perform the classification task. Monte Carlo cross-validation was conducted over 30 folds. A model-agnostic decision-based fusion strategy was adopted for integration. The highest result was obtained by integrating lab data and CT imaging (average AUC= 0.79, p-value< 0.05), while lab data and imaging individually scored average AUCs of 0.77 and 0.7, respectively (p-value< 0.05). Current results showed a modest improvement in the performance using specific complementary modalities over using single modalities. Clinical data did not provide any added value.

Survival Prediction of Cancer Immunotherapy Patients Using Integrative Diagnostics

Abstract:
Introduction: Coronavirus disease 2019 (COVID-19) is an infectious disease caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) (1). On October 9, 2021 authorities have counted 1.261.102 cases and 25.675 deaths in Belgium (2), illustrating the importance of this disease. Reverse Transcriptase Polymerase Chain Reaction (RT-PCR) is considered the gold standard for the diagnosis. The use of chest CT as a screening tool is controversial. Some studies claim that chest CT has a sensitivity of 94.6% and specificity of 46.0% (1) but other studies reported that 10.6% of proven positive patients had normal chest CTs (1). Nevertheless, in the beginning of the pandemic chest CT was widely used as a screening tool partly due to a lack of evidence based diagnostic protocols and availability of RT-PCR tests.
Purpose: To evaluate the performance of an artificial intelligence (AI) tool, for Covid-19 detection and lung injury quantification, on chest CT during a real-time clinical workflow.
Materials & methods: We retrospectively collected a consecutive dataset of Chest CTs obtained between April 1, 2020 and May 1, 2020. These exams were performed to screen for Covid-19 at hospital admission (e.g., emergency, inpatient, and outpatient). All axial images with a slice thickness of 1.5-3 mm were pseudo-anonymized and sent to the AI tool Quibim Precision platform (QUIBIM S.L) (3) to be analyzed by the Imaging COVID-19 Analyzer (This model was developed under the umbrella of a multi-center European research project during the pandemic). The AI tool then provided a probability score for a COVID-19 infection. Chest CTs were evaluated by 12 radiologists with 10-30 years of experience after intramural training on reporting COVID-19 chest CT findings, based on the RSNA expert consensus (4). RT-PCR was considered the gold standard for COVID-19 diagnosis. The concordance between CT findings and RT-PCR was determined by receiver operating characteristic (ROC) analysis and crosstabs.
Results: The study included 264 Chest CTs ( 146 men; 118 women; Mean age: 67,3) with 84 RT-PCR confirmed COVID-19 positive cases, 33 cases of non-COVID-19 related pneumonia and 147 cases without infectious lung disease. When the COVID-19 probability score cut-off value is set at 0.41 there is a sensitivity of 90.48% (95%CI: 82.09% to 95.80%), specificity of 30.00% (95%CI: 23.42% to 37.26%), PPV of 13.65% (95%CI: 12.32% to 15.11%) and NPV of 96.26 (95%CI: 92.77% to 98.10%) and an AUC of 0.75. Regarding the probabilities we suggest the following ranges with 95% sensitivity to exclude the disease and 95% specificity to include the disease: < 0.38: almost certain negative; 0.39 – 0.62: indeterminate; > 0.63: almost certain positive. 
We chose a relatively low cut-off value in order to have a high sensitivity so it could be used as a screening test. However, this reduced the specificity and diagnostic accuracy. When using the suggested probability ranges, a substantial nu

The Performance Of AI Covid-19 Detection And Lung Injury Quantification On Chest CT In A Real-time Clinical Setting

Introduction:
No objective markers to predict remission (absence of depression following treatment) early in antidepressant treatment for patients with major depressive disorder (MDD) have been validated for clinical practice. Structural magnetic resonance imaging (sMRI) can provide valuable information before treatment that may relate to the final antidepressant response; however, the related research findings have been inconsistent.
Hypothesis:
Pretreatment neuroimaging data can be used to predict remission after eight weeks of antidepressant treatment.
Methods:
This study used pretreatment sMRI from a multi-site clinical trial on participants with depression (n=177) who were initiating treatment. The imaging measures were entered into a tree boosting classifier called XGBoost to predict remission after eight weeks, while controlling for age, sex, treatment (antidepressant/placebo), handedness and scan site.
Results:
Our predictive model showed >70% accuracy with 83% specificity for predicting remission, with the 10 most predictive neuroimaging features selected by XGBoost.
Conclusion:
This study was the first effort to predict remission in antidepressant treatment using a tree boosting classifier for neuroimaging data, and provides guidelines to effectively process high dimensional, small sample neuroimaging data for developing predictive algorithms.
Impact:
This study pioneers the application of tree boosting classifier for developing a predictive algorithm using neuroimaging data for antidepressant treatment response.

Early Prediction of Antidepressant Treatment Response Using Machine Learning For Neuroimaging

How Hospitals Need To Think About AI Governance (SIIM)

Q & A - Announce of the Awarded Abstracts - Closing remarks

Teaching Imaging Informatics to Current and Future Generations (SIIM)

[Networking hour](https://app.wonder.me/?spaceId=fe093845-759f-448f-9f55-01d3adccd820)
[COIs](https://drive.google.com/drive/folders/1nbfwAa9lf9Kz-wAZ7iLKhfpBFjpN_gkn?usp=sharing)